FUNCTION OF PANTOTHENATE IN BACTERIA

Vol. 37

425

REFERENCES
Barron, E. S. G. & Lyman, C. M. [1939]. J. biol. Chem.
127, 143.
- Lipton, M. A. & Goldinger, J. M. [1941]. J. biot.
Chem. 141, 957.
Berkman, S., Dorfman, A. & Koser, S. A. [1942]. J. Bact.
43, 6.
Dickens, F. & 8imer, F. [1931]. Biochem. J. 25, 973.
Dixon, M. [1934]. Manometric Methods. Cambridge:
University Press.
Dorfman, A., Berkman, S. & Koser, S. A. [1942]. J. biol.
Chem. 144, 393.
Edson, N. L. [1935]. Biochem. J. 29, 2082.
Fenton, H. J. H. & Jones, H. 0. [1900]. J. chem. Soc. 77, 77.
Fildes, P. [1940]. Brit. J. exp. Path. 21, 67.
& Richardson, G. M. [1937]. Brit. J. exp. Path.
18, 292.
Gladstone, G. P. [1937]. Brit. J. exp. Path. 18, 322.
- [J939]. Brit. J. exp. Path. 20, 189.
Greville, G. D. [1939]. Biochem. J. 33, 718.
Hills, G. M. [1941]. Chem. Ind. 60, 220.
Krebs, H. A. [1937a]. Biochem. J. 31, 661.
[1937 b]. Biochem. J. 31, 2095.
& Eggleston, L. V. [1940]. Biochem. J. 34, 1383.
& Henseleit, A. [1932]. Hoppe-Seyl. Z. 210, 33.
Lipmann, F. [1939]. Nature, Lond., 143, 436.

Lipmann, F. [1941]. Advance8 in Enzymology, 1, 99.

Mellwain, H. [1939]. Brit. J. exp. Path. 20, 330.
Pelezar, M. J. (Jr.) & Porter, J. R. [1940]. Proc. Soc. exp.
Biol., N.Y., 43, 151.
Pilgrim, F. J., Axelrod, A. E. & Elvehjem, C. A. [1942].
J. biol. Chem. 145, 237.
Pratt, E. F. & Williams, R. J. [1939]. J. gen. Phy8iol.
22, 637.
Quastel, J. H. & Webley, D. M. [1941]. Biochem. J. 35, 192.
Smyth, D. H. [1940]. Biochem. J. 34, 1598.
Still, J. L. [1941]. Biochem. J. 35, 380.
Teague, P. C. & Williams, R. J. [1942]. J. gen. Physiol.
25, 777.
Wendel, W. B. [1932]. J. biol. Chem. 94, 717.
Westerkamp, H. [1933]. Biochem. Z. 263, 239.
Williams, R. J. [1941]. Enzymologia, 9, 387.
Mosher, W. A. & Rohrmann, E. [1936]. Biochem. J.
30, 2036.
- Truesdail, J. H., Weinstock, H. H. (Jr.), Rohrmann,
E., Lyman, C. M. & McBurney, C. H. [1938]. J. Amer.
chem. Soc. 60, 2719.
Wohl, A. & Claussner, P. [1907]. Ber. dtsch. chem. Ge8.
40, 2308.
& Oesterlin, C. [1901]. Ber. dt8ch. chem. Ges. 34, 1139.
Woolley, D.-W. [1941]. J. biol. Chem. 140, 311.

Micro-method for Estimating Vitamin A by the Carr-Price Reaction

BY H. HOCH (Freedom Research Scholar), From the Haymead8 Emergency Ho8pital, Bi8hop'8 Stortford,
and the Hale Clinical Laboratory, London Hospital

(Received 16 March 1943)

In the micro-determination here described an attempt has been made to eliminate errors due to
H

LI-*

DCBC A

-~
450mm;. .............-

- 450mm; -

Fig. 1. Photographic arrangement. A, light source, two

100 W. coiled coil bulbs. B, space for filters: (1) 'Calorex'
2 mm.; (2) 'Signal green' 1 mm. (manuf. by -Messrs
Chance, 10 Princes St, S.W. 1). C, diffusing glass 2 mm.
D, reaction tube and standard tubes. E, W-ratten 26 red
gelatin filter. F, enlarging lens at suitable distance to
give image of tubes of about natural size. G, slit, 3-5 mm.
wide, in front of slide which is moved in vertical direction.
H, slide.

fading, differences in the colour of test and standard,

and interference by non-specific absorption. The

tube containing the reaction mixture is photographed in light of a selected wave-length range
together with a series of standardt tubes of equal
dimensions (see Fig. 1),-and the photographic film
is analysed by means of photoelectric photometry.
The advantages of this method are: (1) the working
range is between 0-06 and 0-4 i.u. vitamin A in a
sample; such small amounts cannot be determined
by photoelectric colorimetry in the usual way, but
it is necessary to work in this range when making
serial investigations on the blood of small animals
and is convenient in investigations on human
beings; (2) rapid mixing is ensured as a consequence
of the small volume of the reacting liquids (0.02 ml.);
(3) a record of the colour produced can be obtained
as early as 2 sec. after the initiation of the reaction,
and its course can be followed photographically
with accurate timing, thus avoiding the lag of
galvanometer readings in photoelectric methods
[Dann & -Evelyn, 1938; Yudkin, 1941]; (4) the
readings are independent of fluctuations in the light
source during the reaction.
27-2

426

H. HOCH

A detailed study of the error due to fading in the

Carr-Price reaction is presented below. The principle of the method is generally applicable to
measurement of colour in small volumes.

EXPERIMENTAL
Determination of vitamin A
Standards. Vitamin A. Any stable substance
which has an absorption maximum in the region of
620 m,u is suitable as a standard. Solutions of
cuprammonium sulphate in water were prepared
ranging from 2- 145 to 0<167 g. CUSO4 . 5H20/100 ml.,
decreasing regularly by 20 %. The solutions
contained 6 ml. conc. NH3 (sp. gr. 0.88) and 15 g.
(NH4)2SO4/g. CuSO4. 5H20. The (NH4)2SO4 stabilizes the colour in the presence of traces of alkali
from the glass. Quantities of 0-04 ml. were measured into small glass tubes measuring 2 mm. inside
diameter, 2-9 mm. external diameter and 30 mm. in
length, and these were then sealed. The fused end
was sufficiently flat to give a negligible error in the
height of the columnu of fluid calculated from
volumes exceeding 0;01 ml. The height was measured on a mirror scale and 0 3 mm. was added for
the volume of the meniscus. The average diameter
of 10 tubes cut from one piece of tubing varied less
than 0 5 %; the variation was 1 % in two series
of tubes. In order to eliminate differences in the
diameters of individual tubes correction factors
were applied to the concentrations. The standards
were calibrated with solutions of vitamin Anaphthoate (2-22 i.u./ml. after addition of the
reagent). Original solutions in arachis oil of 1000 i.u.
vitamin A-naphthoate and of 300,g. fl-carotene/g.
were supplied by British Drug Houses, Ltd.
,8-Carotene. Standard tubes were prepared as
described for the vitamin A standards and filled
with solutions of K2Cr2O7, containing 1 ml. conc.
HCI (sp. gr. 1.18)/150 mg. K2Cr2O7, in concentrations from 0 595 to 0-01865 g./100 ml. This standard
matches well with solutions of carotene and extracts
of human and rat blood and rat liver in light
petroleum when compared visually in daylight
without light filters through the range of 1 2-36,ug.
8-carotene/ml. Acid bichromate neutralizes any
alkali which may dissolve from the glass and
its colour is nearer to that of carotene than bichromate. Calibration was carried outwith solutions
of pure fl-carotene in light petroleum (B.P. 40-60').
Fig. 2 shows the calibration curve as obtained by
visual comparison in daylight. The smallest amount
estimated was 0 004,ug. carotene in 0-003 ml.
Procedure. The reaction tube was made by joining a piece
of glass-tubing of a diameter equal to that of the standards
to a wider glass tube of 7 mm. inside diameter, and fusing
the narrow tube at such a distance as to provide a length
of about 15 mm. and in such a way as to give it a flattened

I943

end. The extract freed from water was pipetted into the
reaction tube and the solvent evaporated in a stream of N2
in a bath at 50-55. The wall of the tube was repeatedly
rinsed with small amounts of light petroleum (40-60o) and
the latter evaporated off. After addition of 0-005-0-03 ml.
of light petroleum the tub'e was stoppered and centrifuged.
The volume of the solution was calculated from the height
of the column, allowing 0 3 mm. for the meniscus. The
fl-carotene layer was then mixed with the light petroleum
and the carotene estimated by comparison with the bichromate standards. For many purposes no greater accuracy than 5 % is required and comparison by the eye is
sufficient. Greater accuracy could be obtained by photometry, as will later be described for vitamin A.
The light petroleum was then evaporated in a stream
of N2. With extracts of blood it was sometimes necessary

to transfer this solution from the small amount of precipitate present into another reaction tube with as little
light petroleum as possible (0-04 ml.). The residue was
taken up three times with pure CHC13, evaporated to
dryness to remove all light petroleum, and finally dissolved
in 0-005-0-015 ml. CHC13. Acetic anhydride was added by
capillary pipette to give a final concentration of 1-1-5%
and the tube kept stoppered until the SbCl3 reagent was
added. The acetic anhydride forming the bottom layer was
then mixed with the solution and the reaction tube put in
position as shown in Fig. 1. The SbCl3 reagent (saturated
solution in CHC13 containing 1-5 % ethanol) was added from
a capillary pipette having a long and very fine point. The
flow of reagent was controlled by means of a piece of rubber
tubing attached to the pipette. The point was kept centrally
and near the surface, and the addition occupied about j sec.
The volume of the reagent added was equal to that of the
original solution, and was measured on a mm. scale in the
horizontal position, 0-01 ml. corresponding to 30 mm. The
first exposure of 1I sec. was made 2 sec. after the reagent
was added. Subsequent exposures of 1 sec. were made by
moving the slide rapidly downwards at intervals of 1 sec.
with the camera shutter open. The reaction tube was then
stoppered and the volume of the liquid determined.
Carotene

pg./ml.

0
21

10

0.1

0-2

0-3

0-4

% K,Cr207

Fig. 2.

0-5

0.6
%

CuSO15H20

Fig. 3.

Fig. 2. Calibration curve for carotene.

Fig. 3. Calibtation curve for the Carr-Price reaction of
(a) vitamin A and (b) carotene.
The photographic films were analysed electrophotometrically under a microscope. A disk with a slit parallel
to the photographic images to be compared was attached
to the ocular mounting from which the ocular lens had
been removed and a photoelectric cell of the barrier-layer

Vol. 37

427

ESTIMATION OF VITAMIN A

type was mounted at the upper end of the microscope.

By moving the film under the microscope perpendicularly
to the slit, galvanometer readings were obtained alternately
for maximal transmission through the film (corresponding
to maximal light absorption in a tube) and for the background of the film. The differences between the readings of
the maxima and the background were plotted against the
concentration of CuSO4 in the standards, and the unknown
value was read from the line connecting the values for the
standards. This approximates to a straight line under the
oonditions used (Fig. 2). It is essential that the background
be evenly illuminated. Photometry of three or four adjacent
tubes in the range of tho unknown concentration is sufficient. Fig. 3a shows the calibration line obtained with
vitamin A-naphthoate. The arachis oil containing the

The fading error in the Carr-Price reaction

It is generally accepted that the maximum colour
of the Carr-Price reaction is proportional to the
vitamin A present. Spectrographic examination
confirms this view. Notevarp & Weedon [1936]
showed that the blue colour obeys Beer's law.
McFarlane & Sutherland [1938] achieved excellent
agreement (average error 1 %,) between the two
methods in analyses of fish-liver oil concentrates,
but the error was greater with the-oils themselves
(average error 6%). Jensen & With [1939], ana-

100

5
0

In,
I UU r- 4r--

Ca

unlr

50

c0
03
0

10

Uo

30

20

40 sec. 50

vO )

(a)

100

;-

10

40 sec. 50

30

20
(b)

100

cS

501-

E 50

:0h 0

'4.,0

10
(c,)

20 sec.

10
(d)-

20 sec.

(e)

100

50

c;z
0

K1=0-8

K1 =0'8

K,= 1-2

K2=0-02

K2=0.

K20-05

10
20 sec.
- (h)
(g)
Fig. 4. Fading of Carr-Price colour given by different preparations with vitamin A activity. (a) Vitamin A-naphthoate:
0-283 i.u. (12-9 i.u./ml.). (b) P-Carotene: 0-77/Ag. (35.5,ug./ml.). (c) Blood extract:
(D 0 380 i.u. (16-7 i.u./ml.);
0-135 i.u. vitamin A (5-55 i.u./ml.); 0-143,ug. carotene (5-9jug./ml.). (d) Blood extract with low concentration of
vitamin A and high concentration of carotenoids: 0-054 i.u. vitamin A (2.9 i.u./ml.); 0-337,ug. carotene (16-7,ug./ml.).
(e) Rat-liver extract: 0-264 i.u. vitamin A (13-0 i.u./ml.); 0-041,tg. carotene (2-0,ug./ml.). (f), (g) and (h): calculated
0

10

sec.

10

20 sec.

(f)

fading

curves.

vitamin A-naphthoate was diluted with light petroleum

and aliquots were treated as described for extracts. The
equation of the calibration line was found to be
y = 18.25x

1-03,

x being the concentration of CuSO4.5H20 as a percentage

and y the coilcentration of vitamin A in i.u. The standard
deviation was i 3.3% in the range of 4-22 i.u./ml. With
concentrations approaching 2 i.u./ml., corresponding to
0-04 i.u. in 0-02 ml. of reaction mixture, the error gradually
rises to 15%.

lysing livers of mammals, birds and man, give the

values found for B"S 61/El 328m/h as 2-70, 2;63
anA 2-75 respectively, with a=0-404, 0-436 and
0-406, where B1 %. S61 represents the extinction
coefficient of the Carr-Price reaction mixture (light
filter S 61) and El % 328 the extinction coefficient
for light of wave length 32AmtL.
Variable fading rates have been observed by
Dann & Evelyn [1938] not to affect the maximum
colour intensity; and Meunier & Raoul [1938]

H. HOCH

428

suggest that they may serve as additional criteria

for distinguishing vitamin A from vitamin A2.
No information could be found concerning colour
development during the first j-min. after the start
of the SbCl3 reaction. This was studied by the
method described above in samples of pure vitamin
A-naphthoate and ,B-carotene, and in extracts of
human blood and rat liver. If the maximum colour
is reached before fading starts, these curves (Fig. 4)
indicate a variable onset of the latter and variable
rates of fading. If, however, fading begins as soon
as the coloure-d compound is present, i.e. at the
time of adding the reagent, the ideal maximum is
never reached, but its value could be calculated by
applying a correction, derived from the fading rate,
to the observed maximum.
The calculations are based on a first order type
reaction curve both for colour development and for
fading in the neighbourhood of the maximum. The
fading rate is assumed to be proportional to the
concentrations of the molecules of vitamin A which
had reacted with the SbCl3, giving the blue colour.
The relation between observed blue colour, expressed as percentage of the maximum colour assuming that no fading has taken place, and the
time, is given by
kk (e-kit _k2t
1OOx
t
kl
aO
k2.the solution of the differential equation
dx/dt = aokle kLt _ k2x;
where x = concentration of vitamin A as chromogenic compound, ao = concentration of vitamin A at
the beginning of the reaction, k, = reaction constant
of the colour development, k2 = reaction constant
of the fading reaction, t = time in sec.
The values of the ordinates corresponding to a
series of finite exposures as used in these experiments (1j and 1 sec.) were calculated from
10 t2'xdt
12
1 / 1
aO
+ k2
h_tj k2-kl kItL
Fig. 4 shows curves (f, g and h) calculated by means
of this formula with varying values for k, and k2,
in which the ordinates represent the values for
100x/ao expressed as percentage of the highest value
for lOOx/ao obtained.
It is possible to estimate k2 directly from the
slopes of the experimental curves at their points of
inflexion after the maximum, when colour formation has practically stopped. From this figure and
the value of the ordinate obtained at the first
reading, k, can be determined by the formula, and
with it, theoretically, the time of the true maximum.
k2 was 0-018 0-003 for vitamin A-naphthoate
and 0-005 for carotene. Wide variation from 0-003
to 0-1 was observed with extracts from various
-

I943

sources. The value of the first ordinate in 55 experiments ranged from 90 to 100% of the maximum
value found. The accuracy of the determination of k1
was limited by the practical difficulty of ensuring
complete mixing within 2 sec.
When k, is taken as 0-7 and k2 as 0-01, the first
ordinate works out at 88-5 % of the maximum.
Where kL is 0-7, or greater, and k2 is 001, or greater,
the position of the maximum according to the above
formula must be at or earlier than 6 sec. after
reaction. The positions of all maxima were found
to fulfil this condition.
In dilutions below 4 i.u. vitamin A/ml. the curves
often deviate from the calculated, the fading constant being greater at the beginning than would
correspond to the value found at 22 sec. after
reaction. The effect was not detected with concentrations above 4 i.u.Iml. or with solutions of
vitamin A-naphthoate. It may be caused by interfering substances in the blood. Agreement is obtained between curves calculated with k, = 0-8 and
experimental curves for concentrations greater than
4 i.u./ml., but the validity of the formula derived
needs further test on accoAunt of theuncertainty of kc .

Table 1. The estimation of vitamin A and carotenoid8 and the fading rate of the Carr-Price colour
with two concentrations
Conc. in blood
per 100 ml.

Subject
I.V.
M.

B.
A.B.

J.R.M.
R.B.
D.M.

J.R.M.
W.

serum or
% fading
for
plasma
Serum
Conc. in
A
or
vit. A at
reaction(mixture
pg. plasma 22 sec.
after
i.u. vit. i.u. caro- used
ml. reaction
A/ml. vit. A tene
45
51
2-1
0-081
57
0-156
49
55
33
3-6
32
1-24
25
0 09

Vol. of
reaction
mixture
mi.
0-0199
0-0213
0-0184
2-59
0-0195
2-67
0-0177
4*50
0-0189
4-37
'0-019
7'52
0-0213
5-27
0-0187
7-24
0-0216
2-92
0-021
6-3
0-0195
6.52
0-0187
0-0186 12-7
4-12
0-021
7.55
0-0196
2-66
0-0177
3-64
0-021

28

31

53
47
103
103
97
87
76
78
135
131
85
82
52
43

64
73
90
90
78
83
97
95
87
89
90
89
74
78

0-18
0 09

0-18

0-081
0-156
0-102
0-18
0081

24-3
18-2
5.5
8
43

0-156
0 09
0-18
0-102
0-18

32

0-09

11-7
14-5

0*18

The observation that the percentage fading measured by the value at 22 sec. in two different concentrations of an individual extract is nearly the
same (Table 1), may be taken as evidence of a first
order type of fading curve. Whether the mechanism
of the fading reaction is unimolecular or whether
it is bimolecular with a second constituent in excess
was not investigated.

Vol. 37

ESTIMATION OF VITAMIN A

It is of practical interest to know-whether corrections shoVld be applied to the maximum colour

value observed in samples that fade at different
rates. Table 2 gives the calculated maximum values

429

that described for vitanxin A (Fig. 3b). Values for

concentrations lower than 8,ug. carotene/ml. were
too small to be measured accurately. For the purpose of correcting the Carr-Price values of extracts

Table 2. Calculated correction8 to maxima, reached with different value8 of k, and k2

k2 0

k1
0-8
1*0

005

0*03

0.01
0002

0050

0.1

Max. Corr. f22 Max. Corr. f22 Max. Corr. f22 Max. Corr. f22 Max. Corr. f22 Max. Corr. f22
97*2 - 6 5.1 94.5 -3.5 14*0 90 9 +0*5 27-5 87.6 +4 38-7 82*9 +10 57-2 73-4 +24.5 82*7
97-7 -5 5-9 95*3 -2-5 15.1 92-3 +05 28-9 89-4 +4 404 84-5 +10 58-5 76X6 +21.5 83.9
Max. = calculated maximum value of colour as percentage of theoretical total, without fading.
Corr. =corrections to experimental maximum values as percentage.
f22 =fading 22 sec. after reaction as percentage of the maximum.

as percentages of the theoretical values without

fading (1st columns), and the corresponding corrections for samples whose constants of fading differ
from those of vitamin A-naphthoate (2nd columns).
The 3rd columns give the percentages of fading
22 sec. after reaction.
The distribution of the constants of fading in
33 blood extracts is given below. These figures were
calculated from-the values for the blue colour due
to the vitamin A,. and it was assumed, though not
quite correctly, that the yellow colour of the extracts was entirely due to ,B-carotene. Possible differences in fading rate of the blue colour formed with
fl-carotene with various extracts were disregarded:
0-01- 0-02- 0-03- > 0-04
k2 < 0-005 0-0050-03
0-01
0-02
0-04
No. of
2
2
4
11. 14
'0

samples

'

The corrections for the majority of extracts from

blood are within + 5 % (see Table 2), and so they
need not be applied for practical purposes, but with
fading rates greater than k2 = 0-05 the observed
maxima yield values which are appreciably too
low, and corrections are necessary.

The correctionfor p-carotene in the Carr-Price reaction

The blue colour given by carotene with the CarrPrice reagent was measured in the same way as

containing carotene the calibration line was extrapolated graphically to reach the point where the
calibration line for vitamin A crosses the abscissa,
and this value, 0-056 % CuS04. 5H20, was subtracted from all figures of the calibration curve for
carotene. The reason why the calibration line for
vitamin A does not reach the zero point may be a
smaller light transmission of the control compared
with that .of water.

SUMMARY
1. A photographic device for the measurement
of the Carr-Price reaction is described.
2. The rates of fading with different extracts
were studied, and the corrections which it is necessary to apply to the values obtained were determined.
3. If the concentration of vitamin A in the blood
is calculated from the observed maximum colour'
without correction for fading the error will generally
not be greater than 5 I wish to thank the authorities of the Haymeads Emergency Hospital, Bishop's Stortford, and of the London
Hospital for the facilities which made this work possible and
the British Drug Houses, Ltd. for preparations of vitamin A
and carotene.

REFERENCES,
Dann, W. J. & Evelyn, K. A. [1938]. Biochem. J. 32,
1008.
Jensen, H. B. & With, T. K. [1939]. Biochem. J. 33, 1771.
McFarlane, W. D. & Sutherland, A. J. [1938]. Canad. J.
Re8. B, 16, 421.

Meunier, P. & Raoul, Y. [1938]. C.R. Acad. Sci., Pari8,

206, 1148.
Notevarp, 0. & Weedon, H. W. [1936]. Biochem. J.
30, 1705.
Yudkin, S. [1941]. Biochem. J. 35, 551.
--